Every smartphone in today’s world is most likely powered by an ARM-based processor. But, the chipmaker is now looking to push the envelope by banking heavily on artificial intelligence for the next-generation of its mobile processors. With regards to the same, ARM is today unveiling its latest multi-core architecture technology called DynamIQ.
This new multi-core microarchitecture will serve as the foundation for the chipmaker’s upcoming Cortex-A processors, which will be efficient in handling computing demands for AI software solutions. ARM has already been using a big.LITTLE architecture, which is also a multi-core architecture that allows phone makers to couple a low-powered chip with a high-powered one to extend battery life without losing on processing power. But, this has now been extended with their DynamIQ technology, which provides heterogeneous processing capabilities at the compute model itself.
Here, Softbank-owned ARM means to suggest that the new central processing unit (CPU) architecture will now enable chipmakers to cluster even more, up to eight, heterogeneous processor cores onto one chip. This continues to follow the big.LITTLE ideology, where each core can be tailored to perform tasks as per the corresponding software. Each core of the chip can have different performance, as well as power characteristics. It also provides faster data access and enhanced power management because of a redesigned memory subsystem.
This means that the CPU includes an exclusive core for handling AI operations, which will be accompanied by software libraries to run the most popular ones. ARM boasts that Cortex-A processors designed on its DynamIQ architecture will boost artificial intelligent and machine learning performance 50 times in the next 3 to 5 years. ARM’s partners are already working on DynamIQ technology, that’ll be an optimal, responsive and effective powerhouse for future devices.
Talking about the future of DynamIQ technology in an official release, Nandan Nayampally, general manager of the ARM Compute Products Group said,
The flexibility and versatility of DynamIQ will redefine the multi-core experience across a greater range of devices from edge to cloud across a secure, common platform.
DynamIQ technology will be pervasive in our cars, our homes, and of course our smartphones as well as countless other connected devices where machine learning is applied to the zettabytes of data they generate — both within the cloud and at the device level.
In addition, ARM continues to boast that the DynamIQ technology is the biggest shift in architecture that the company has launched since big.LITTLE in 2011. The latter has been implemented across some of the most used chips, such as Qualcomm’s Snapdragon for Android and Apple’s A10 Fusion for iPhones. And now, the U.K-based chipmaker expects the adoption of DynamIQ to be even more far-reaching due to exploding scope for artificial intelligence solutions in the tech space.
While most of the current AI and ML applications are powered by the cloud, ARM is planning to further the development by bringing processing closer to the compute board. With the advent of AI technology in almost every technological ecosystem, ARM is trying to push its technology to even more avenues beyond smartphones to autonomous cars, and the wide range of IoT devices. This has now become particularly necessary as processing needs to happen instantly and poor cellular connectivity could disrupt the experience. It is, thus, increasingly important to handle processing tasks locally at the processor within the device.
ARM further adds that their new processor series, dubbed as Cortex-A, will be based on the DynamIQ architecture. These chips, along with its partners, will be announced later this year. And since they’re no longer restricted to smartphones, so you could expect devices based on this technology to ship sometime early next year i.e in 2018. The chipmaker recently achieved the 100 billion milestone and is now looking to ship about 100 billion more processors by 2021.
